Well damn. Amusing how sometimes something can be both funny and sad.

Holy shit really? Do you have a source for this?

Working title: “Sloppy Seconds”.

Borderline strawman there but I’ll bite.

Open weight models trained unethically are unethical. Closed weight models trained unethically and then sold back to you for a profit from gas-powered datacenters funded through Ponzi schemes are substantially more unethical.

From there it’s a harm reduction calculus. No, those are never pleasant.

So do you let the closed weight labs conquer the field unopposed just so you can feel better about yourself? That’s a valid stance, FWIW, and it’s also super easy and convenient because you don’t have to do anything. It especially makes sense if you believe it’s still possible that AI will just go away on its own. I don’t, myself, not anymore, so I encourage the use of open weight models, however grudging, so people don’t give money to the closed labs and in the worst case aren’t eventually stuck with the maximally unethical options. And we’ve not even touched on the nightmare labor replacement scenarios that seem every day less unlikely. Am I right? I have no clue. Like you, I’m just trying to make the best choices I can in a world that’s gone to shit. I’d recommend dropping holier-than-thou attitude either way, though, because it doesn’t help our side. Man.

You run a program like llama.cpp that can use the weights to run the model, which you can then use for whatever you’d use a model for: figuring out tech stuff, coding, etc. It’s a bit involved, but there are tools that make it easy to get started. LM Studio for instance.

In some cases, the lab that created the model publishes the dataset that it was trained on, and those are usually made of publicly available data. In most cases, though, the labs don’t give details, but the answer likely involves siphoning every web page they could find.

It means the weights (the numeric data that makes up the model) are publicly available to download and use for free.

In some cases there are conditions such as, if you run the model as your business, you need to purchase a license. That’s usually the largest, most powerful models, though, not those most people would be able to run at home.

Fuck all the war songs, Slovenia’s anthem is a drinking song called The Toast if I remember correctly.

Hanggai also kicks ass, in the rarefied yet delightful niche of Mongolian metal.

That’s a question for the ages for sure. I’m worried it’s going to get quite a bit worse before it gets better.

As far as I can tell, there are multiple ongoing components to what’s going on.

There is definitely some Russia-driven agitprop going on, with the apparent aim of weakening the West through division. But on its own I think it’s not a sufficient explanation, for two reasons. The same research showed that Russia is also pushing left-aligned agitprop points to increase division. And Russia was not similarly involved in the 1930’s, which nevertheless saw a somewhat similar trajectory.

But it’s still a factor either way of course.

The main force, I think, is the collapse of capitalism. That capitalism is its own self-destruction mechanism has been observed for centuries; it’s kind of one of the main takeaways from Das Kapital. With the ceaseless rise of wealth capture by the ruling classes, discontent rises correspondingly among the general population, and the beige centrist neoliberals that tend to occupy power seats in most democracies keep onboarding more and more far right talking points in an apparent effort to keep this discontent from getting an actual left wing government elected: drawing false equivalency between left and far right policies, forming coalitions with the far right to prevent even a mildly left wing party from forming a government, etc.

The latter seems particularly driven by the flow of donor money from the wealthy elites.

But it also looks like in many countries, the national media are disproportionately captured by said wealthy elite, who also increasingly push the same far right societal framing for the same core reason.

Historically speaking, this private capture of wealth and power does not stop on its own, and it goes to some pretty bad places before it gets better. It’s bleak as hell, but we can take hope, I think, from another historical observation: it always gets better, eventually.

Till then, take care of each other, y’all.

From what I can observe, yeah, pretty much the entire industry has shifted to LLM code generation. I’m sure (or at least I hope) there are still some shops that go out of their way to keep working like we did, you know, months ago. But beyond that… I’m not even sure the general public understands how quickly and how fully LLM codegen has become the norm. It’s grim.

There’s nothing to prevent you from writing code in an IDE while the torment nexus TUI churns out whatever monster misfeature request was waiting for you in your Jira queue this morning.

In fact I’d recommend doing the code writing part. It helps, even if it increasingly becomes a rounding error in your overall output.

It’s bleak as shit right now. God why couldn’t this have waited until after my retirement.

I understand Costco, I think it was, consistently made the opposite choices from Target. That’s off the top of my mind, though, so do check for yourself.

Dang, I only spotted your comment after posting the same joke. Here, have an upvote instead.

deleted by creator

Yeah, Minus is basically magical (and more than a little autism-coded) in a world without magic. It’s a strange comic, but in a good way.

Well crap. Sorry too hear. I hope it does go away in time. Long COVID is no joke.

The Internet was already a very different place back then. These days, most people are on a few very large platforms. Back then it was a lot of smaller sites and forums, whose individual vibes depended a lot on its own moderation policies.

I spent much of that day with other players of the MUD-style game I played at the time, figuratively huddling together, sharing updates about our people in NYC as we heard from them.

For a time, I think, people were kind of stunned. The US had thought itself invulnerable, you understand.

But it wasn’t very long before the raging islamophobia began, and a hard turn toward the direction of fascism. But also the reaction to that. The split was stark.

And is still going on to this day, I suppose.

So no, people didn’t quite stop being dicks. If anything they only started dicking harder.

Interesting point… In this case, I have a hunch that the requirements in terms of training data and hardware to run the training on would remain prohibitive, but I do wonder how much could be actually achieved using a distributed scheme of sorts.

I’d take a model whose training data is open source and legitimately obtained. The only ones I know about are Apertus and OLMo, and they aren’t really competitive.

Yeah, I don’t actually know what’s up for the Gulf of Mexico but it does seem they implemented a mechanism to show two names in certain contexts. I’ll theorize it triggers when a polygon overlaps international jurisdictions with conflicting data, but that’s just a guess and therefore likely wrong.

At any rate there are probably intense internal meetings going on right now…